LIDAR systems and methods for detection and classification of objects

Abstract

A vehicle-assistance system for classifying objects in a vehicle's surroundings. The system includes: at least one memory configured to store classification information for classifying a plurality of objects; and at least one processor configured to receive a plurality of detection results associated with light detection and ranging system (LIDAR) detection results, each detection result including location information, and further information indicative of at least two of the following detection characteristics: object surface reflectivity; temporal spreading of signal reflected from the object; object surface physical composition; ambient illumination measured at a LIDAR dead time; difference in detection information from a previous frame; and confidence level associated with another detection characteristic. The at least one processor is also configured to: access the classification information; and based on the classification information and the detection results, classify an object in the vehicle's surroundings.

What is claimed is: 
     
       1. A vehicle-assistance system for classifying objects in a vehicle's surroundings, the system comprising:
 at least one memory configured to store classification information for classifying a plurality of objects; 
 at least one processor configured to:
 receive a plurality of detection results associated with light detection and ranging (LIDAR) detection results, each detection result including location information, and further information indicative of at least two of the following detection characteristics:
 object surface reflectivity; 
 temporal spreading of signal reflected from the object; 
 ambient illumination measured at a LIDAR dead time; 
 difference in detection information from a previous frame; and 
 confidence level associated with another detection characteristic; 
 
 access the classification information; and 
 based on the classification information and the detection results, classify an object in the vehicle's surroundings. 
 
 
     
     
       2. The vehicle-assistance system of  claim 1 , wherein each detection result including location information, and further information indicative of at least three of the following detection characteristics:
 object surface reflectivity; 
 temporal spreading of signal reflected from the object; 
 object surface physical composition; 
 ambient illumination measured at a LIDAR dead time; 
 difference in detection information from a previous frame; and 
 confidence level associated with another detection characteristic. 
 
     
     
       3. The vehicle-assistance system of  claim 1 , wherein the at least one processor is configured to classify a first object as a car based on the classification information and on a plurality of first detections results; to classify a second object as a road based on the classification information and on a plurality of second detections results; and to classify a third object as a person based on the classification information and on a plurality of third detections results. 
     
     
       4. A computer-implemented method for object classification, comprising executing on a processor the steps of:
 processing first light detection and ranging (LIDAR) detection information to provide a first three dimensional (3D) model; 
 processing second LIDAR detection information, obtained after the first LIDAR detection information, to provide a second 3D model; 
 identify a first object detection in the first 3D model; 
 identify a second object detection in the second 3D model; and 
 based on the first object detection and the second object detection, determine a classification for an object detected by the LIDAR. 
 
     
     
       5. A non-transitory computer-readable medium for object classification, comprising instructions stored thereon, that when executed on a processor, perform the steps of:
 processing first light detection and ranging (LIDAR) detection information to provide a first three dimensional (3D) model; 
 processing second LIDAR detection information, obtained after the first LIDAR detection information, to provide a second 3D model; 
 identify a first object detection in the first 3D model; 
 identify a second object detection in the second 3D model; and 
 based on the first object detection and the second object detection, determine a classification for an object detected by the LIDAR. 
 
     
     
       6. The vehicle-assistance system of  claim 1 , wherein the classification information includes a database of known fingerprints including at least one of reflectivity fingerprints, surface angle fingerprints, or point cloud fingerprints. 
     
     
       7. The vehicle-assistance system of  claim 1 , wherein the detection results include confidence levels associated with the detection characteristics, and
 wherein classifying the object in the vehicle's surroundings includes, based on the classification information and the confidence level associated with each detection result, classifying a plurality of pixels as being associated with the object. 
 
     
     
       8. The vehicle-assistance system of  claim 1 , wherein the classification information includes known fingerprints including at least one of reflectivity fingerprints, surface angle fingerprints, or point cloud fingerprints, and
 wherein classifying the plurality of pixels as being associated with the object includes comparing the detected results with the known fingerprints included in the classification information to identify a match. 
 
     
     
       9. The vehicle-assistance system of  claim 8 , wherein comparing the detected results with the known fingerprints included in the classification information to identify the match includes:
 using one or more neural networks to identify the match. 
 
     
     
       10. The vehicle-assistance system of  claim 1 , wherein the at least one processor is further configured to determine, based on the detection results, at least one of a shape of the object, a spatial relationship between multiple portions of the object, or a reflectivity relationship between the multiple portions of the object, and
 wherein classifying the object in the vehicle's surroundings includes classifying the object based at least in part on the shape of the object, the spatial relationship between the multiple portions of the object, or the reflectivity relationship between the multiple portions of the object. 
 
     
     
       11. The vehicle-assistance system of  claim 1 , wherein the detection results include a velocity of the object, and
 wherein classifying the object in the vehicle's surroundings includes classifying the object based on the velocity of the object. 
 
     
     
       12. The vehicle-assistance system of  claim 1 , wherein the at least one processor is further configured to use a plurality of cascading classifiers to classify the object. 
     
     
       13. The vehicle-assistance system of  claim 1 , wherein the at least one processor is further configured to:
 based on the classification result, determine that the object is an object of interest; and 
 forward, to a LIDAR system which generates the LIDAR detection result, an indication of a region of interest that includes the object. 
 
     
     
       14. The vehicle-assistance system of  claim 1 , wherein the at least one processor is further configured to:
 when a confidential level associated with a detection characteristic is under a threshold, transmit a request for additional information to a LIDAR system. 
 
     
     
       15. The vehicle-assistance system of  claim 1 , wherein the at least one processor is further configured to:
 cause a change in an operational state of the vehicle based on the classification of the object. 
 
     
     
       16. A computer-implemented method for object classification, the computer-implemented method comprising:
 receiving a plurality of detection results associated with light detection and ranging system (LIDAR) detection results, each detection result including location information, and further information indicative of at least two of the following detection characteristics: 
 object surface reflectivity; 
 temporal spreading of signal reflected from the object; 
 ambient illumination measured at a LIDAR dead time; 
 difference in detection information from a previous frame; and 
 confidence level associated with another detection characteristic; 
 accessing classification information stored in at least one memory, the classification information including classifications for a plurality of objects; and 
 based on the classification information and the detection results, classifying an object in the vehicle's surroundings. 
 
     
     
       17. A non-transitory computer-readable medium storing instructions executable by at least one processor to perform a method, the method comprising:
 receiving a plurality of detection results associated with light detection and ranging system (LIDAR) detection results, each detection result including location information, and further information indicative of at least two of the following detection characteristics: 
 object surface reflectivity; 
 temporal spreading of signal reflected from the object; 
 ambient illumination measured at a LIDAR dead time; 
 difference in detection information from a previous frame; and 
 confidence level associated with another detection characteristic; 
 accessing classification information stored in at least one memory, the classification information including classifications for a plurality of objects; and 
 based on the classification information and the detection results, classifying an object in the vehicle's surroundings. 
 
     
     
       18. The computer-implemented method of  claim 4 , further comprising:
 determining a velocity and a moving direction of the object based on a comparison between the first object detection and the second object detection, 
 wherein determining the classification for the object includes determining the classification based on the velocity and the moving direction of the object. 
 
     
     
       19. The computer-implemented method of  claim 4 , wherein each one of the first 3D model and the second 3D model is a point cloud model. 
     
     
       20. The computer-implemented method of  claim 4 , further comprising:
 processing at least one of the first object detection information or the second object detection information for assessing a speed of the object based on at least one of a smear or a blurring resulting from movement of the object during the capturing of the corresponding LIDAR detection information, 
 wherein determining the classification for the object includes determining the classification based on the speed of the object.